# facebookresearch/moco-v3

PyTorch implementation of MoCo v3 https//arxiv.org/abs/2104.02057

Repository: https://github.com/facebookresearch/moco-v3
Canonical: https://ross.abutalabs.com/products/moco-v3
Language: Python
License: NOASSERTION
License Family: other
Archived: true
Last push: 2021-11-25T09:33:38+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1904, "days_push": 1742, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1323, forks 174 (observed 2026-08-28T04:04:22.227625+00:00)

## What it is
A PyTorch implementation of MoCo v3, a self-supervised contrastive learning method for ResNet and Vision Transformer (ViT) models. It includes training code, configs, and pre-trained models based on ImageNet-1k pre-training.

## Use cases
- pretrain vision transformers with self-supervised learning
- reproduce MoCo v3 paper results on ImageNet
- get pretrained ResNet or ViT backbones without labels
- run contrastive pretraining on GPUs instead of TPUs
- evaluate self-supervised models with linear probing or fine-tuning

## When to choose
- you need self-supervised pretrained ViT or ResNet backbones
- you want to reproduce or build on MoCo v3 research
- you have large-scale GPU resources for ImageNet-scale pretraining

## When to avoid
- you need a maintained production library rather than research code
- you lack multi-GPU resources for large-batch contrastive training
- you need self-supervised learning for non-image modalities

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training
- domain: machine-learning, deep-learning, computer-vision, image-processing
- platform: python
- tags: self-supervised-learning, vision-transformer, resnet, pytorch, imagenet, contrastive-learning, pretrained-models, research-code, gpu, linux

## Member repositories
- facebookresearch/moco-v3 (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:22.227625+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:47:04.563811+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/moco-v3 (fetched 2026-08-28T04:04:22.227625+00:00, sha 1a9b04696764)
- Data as of 2026-08-30T08:39:29.467469+00:00.